Anchoring

In a mall, a coat's tag reads "Was 1,299, now 599." Feels like a bargain. If the tag only said "599," would it still seem cheap?

The 1,299 is an anchor: it gets into your head first, and everything afterwards revolves around it.

Even random numbers can anchor you

In a classic experiment, psychologists spun a "wheel of fortune" rigged to stop at 10 or 65, then asked people to estimate the percentage of UN member countries that are in Africa. People who saw 10 gave a median estimate of 25%; those who saw 65 said 45%. Even knowing the wheel had nothing to do with the question, the number still pulled their estimates.

Why? We tend to start from a reference point and adjust toward what seems reasonable, but we usually don't adjust far enough.

Common anchors in forecasting

• Other people's numbers: an expert on the news says "70% likely," a friend says "definitely."
• Your own previous probability: reluctance to leave your old number when updating (which also causes the underreaction from lesson 4.2).
• Numbers in the question: "Will this film gross over 3 billion yuan?" That 3 billion quietly shapes your box-office estimate.
• Whatever first pops into your head.

Below, look at a random number first and then make an estimate, and experience anchoring for yourself.

How to guard against it when forecasting

• Write your base rate first: before looking at anyone else's numbers, start from a reference class and form your own estimate (see lessons 2.1 and 2.2).
• Try several starting points: estimate once starting high and once starting low, and see how far apart you end up.
• Ask: "If I'd just seen a different number, would my estimate be different?"

How to spot it in a review

Look back at your reasoning: was your probability suspiciously close to some outside number (an expert's forecast, a threshold in the question, your last probability), without you being able to say why it was exactly that? Then you may have been anchored.